Project Summary Integration of sensory signals is crucial for sensory processing in the brain. Rodent whisker somatosensory cortex (S1) provides a powerful system to study this, because neurons integrate tactile information across a discrete array of whiskers. S1 circuitry is well characterized, but tactile representation remains poorly understood. Most studies have focused S1 encoding of single-whisker deflections. This stimuli elicit low-probability spiking, relatively weak somatotopic tuning, and highly similar receptive fields and maps in thalamus and across cortical layers. This weak- redundant code suggests that S1 neurons may code for more complex stimuli than single-whisker deflections. My project tests whether S1 neurons code for multi-whisker stimuli, which are generated during natural whisker sensation. I propose that multi-whisker tuning is achieved by linear and non-linear integration and that multi- whisker features are represented in a novel topographic map in S1. Aim 1 describes past work in which I extensively characterized multi-whisker tuning to 2-whisker sequences, which represent a tractable and important subset of multi-whisker stimuli. I discovered that many neurons have strong spatiotemporal tuning for specific 2- whisker sequences at specific inter-whisker-deflection-intervals (?t). I found that a combination of linear and nonlinear mechanisms construct and enhance spatial selectivity, with prominent sublinear suppression of non-preferred stimuli. I also discovered principles governing ?t tuning and that it enhances spatial selectivity for 2-whisker sequences, thus defining general computations underlying multi-whisker integration In Aim2, I will use 2-photon Ca2+ imaging to determine the representation of 2-whisker sequences at the population level. While neurons in L2/3 of S1 are highly intermixed by single-whisker tuning (Sato et al. 2007, Clancy KB et al. 2015), I propose that discrete cortical columns will be apparent if receptive fields are defined in terms of 2-whisker sequences. My results suggest that the edges of each column will be defined by a ring of spatiotemporally selective neurons that form discrete borders between columns through sharp differences in ?t tuning. Also, by imaging many neurons simultaneously, I will test whether firing correlations between spatiotemporally selective and non-selective units provide a robust population level code for 2-whisker sequences. This will greatly strengthen our understanding of how S1 represents multi-whisker stimuli. In Aim 3, I plan to extend my study of active sensory systems by studying the role of motor circuits in perception during active sensation. Overall, my research plan will contribute significantly to understanding how the brain integrates sensory and motor signals to generate accurate percepts of the world.